Search

We aimed to determine the frequency of human papillomavirus-independent (HPVI) cervical squamous cell carcinoma (SCC) and to describe clinicopathologic characteristics. Among 670 patients with surgically treated SCCs in an established multi-institutional cohort, 447 had available tissue. Tissue microarrays were constructed and studied by in situ hybridization (ISH) for high-risk and low-risk human papillomavirus (HPV) mRNA and immunohistochemistry for p16 and p53. Tumors were HPVI if negative by HPV ISH and they failed to show diffuse p16 positivity by immunohistochemistry, and human papillomavirus-associated (HPVA) if positive by HPV ISH. Ten HPVI SCCs and 435 HPVA SCCs were identified; 2 cases were equivocal and excluded from analysis. The overall rate of HPVI SCC was low (2%) but was higher among older patients (7% in patients above 60 y of age and 17% in patients above 70 y of age). Compared with HPVA, patients with HPVI SCC were significantly older (median age, 72 vs. 49, P <0.001) and diagnosed at a higher stage (40% vs. 18% with stage III/IV disease, P =0.055). p53 expression was varied; 2 cases (20%) had null expression and 8 (80%) had wild-type expression. HPVI SCCs were heterogenous, with keratinizing, nonkeratinizing, and warty morphologies observed. Several cases had a precursor lesion reminiscent of differentiated vulvar intraepithelial neoplasia, with prominent basal atypia and hypereosinophilia or a basaloid-like morphology. Two patients (20%) had distant recurrences within 12 months, and 3 (30%) died of disease during follow-up. HPVI SCCs are rare tumors that are more common among older patients with higher stage disease and have important clinical and histologic differences from HPVA SCCs., Competing Interests: Conflicts of Interest and Source of Funding: Funded in part by the National Institutes of Health (NIH)/National Cancer Institute (NCI) Cancer Center Support Grant P30 CA008748. Outside the submitted work, N.R.A.-R. reports research funding paid to the institution from GRAIL. A.I. reports consulting fees from Mylan. For the remaining authors none were declared., (Copyright © 2023 Wolters Kluwer Health, Inc. All rights reserved.)

HIK1083 and trefoil factor 2 (TFF2) are known to be expressed in gastric-type carcinoma (GAS), but they do not reliably mark all GASs, and focal expression can be missed in biopsy specimens. We aimed to investigate whether claudin-18 and alpha-methylacyl-CoA racemase (AMACR) could be surrogate markers to separate GAS from other types of endocervical adenocarcinoma (ECA) and to compare their usefulness with that of HIK1083 and TFF2. Claudin-18 and AMACR immunohistochemistry was performed, and the results were compared with that of TFF2 and HIK1083, using whole sections of 75 ECAs (22 GASs and 53 non-GASs) and 179 ECAs with tissue microarrays (TMAs). TMAs were built to simulate the assessment of immunohistochemical stains in small biopsies. Any membranous (claudin-18) or cytoplasmic/membranous (AMACR, TFF2, HIK1083) staining of >5% of tumor cells was considered positive. Of 75 ECAs with whole sections, claudin-18 was significantly more frequently expressed in GASs (21/22) compared with non-GASs (8/53) (P<0.01). In ECAs with TMAs, claudin-18 expression was significantly frequent in GASs (15/23, 65.2%) than in non-GASs (3/152, 2.0%; all usual-type) (P<0.01). All claudin-18-positive GASs showed intense staining except 1 case. Claudin-18 shared the same degree of sensitivity and specificity with HIK1083 and TFF2. Three clear cell carcinomas were positive for claudin-18, but none showed intense staining. AMACR was expressed in a subset of ECAs and showed no impact in distinguishing between GAS and other ECAs. Our results suggest that claudin-18 is a promising surrogate marker to separate GAS from other types of ECA, including clear cell carcinoma., Competing Interests: Conflicts of Interest and Source of Funding: Supported in part through the NIH/NCI Cancer Center Support Grant P30 CA008748 (R.A.S., K.J.P.). The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.)

The International Endocervical Adenocarcinoma Criteria and Classification (IECC) categorizes endocervical adenocarcinomas (ECAs) on the basis of morphologic features linked to etiology (ie, human papilloma virus [HPV] infection), resulting in separation of ECAs into HPV-associated (HPVA) and unassociated or non-HPVA (NHPVA) types. NHPVAs are reported to be large and present at high stage in older individuals. Our aim was to examine the clinical outcomes in these tumor types. Full slide sets of 205 ECAs were collected from 7 institutions worldwide and classified on the basis of IECC criteria and the presence or absence of HPV. Clinical and morphologic parameters were correlated with follow-up data. Statistical analysis of overall survival (OS), disease-free survival (DFS), and progression-free survival (PFS) were conducted using the Kaplan-Meier survival analysis and compared using the log-rank test for univariate analysis. Multivariate survival analysis was conducted, and the survival endpoints considered were OS, DFS, and PFS. Statistically significant survival differences (OS, DFS, and PFS) were found when comparing the following categories: HPVA>NHPVA (ie, survival was superior in the setting of HPVAs), including patients treated with surgery followed by adjuvant therapy; usual-type HPVA>mucinous HPVA; FIGO grade 3 HPVA>NHPVA; HPVA>NHPVA, both with lymphovascular invasion; and HPVA>NHPVA in patients with pelvic recurrences. Although there were trends favoring HPVA outcomes over those of NHPVA, these differences were not statistically significant in the following categories: mucinous HPVA versus NHPVA; HPVA versus NHPVA, both with lymph node metastases at presentation; and HPVA versus NHPVA in patients with distant metastasis. Survival for both HPVA and NHPVA was similar when surgery without adjuvant therapy was used. FIGO grading did not have prognostic significance in HPVAs. Multivariable analysis of HPVAs indicated nearly significant statistical associations between stage and both OS and DFS (P=0.07 and 0.06, respectively), and between Silva invasion pattern and OS (P=0.09). Multivariate analysis of NHPVAs indicated a statistically significant association between OS and age (P=0.03), stage (P=0.02) and tumor size (P=0.002), and between DFS and stage (P=0.004) and tumor size (P=0.004). Multivariate analysis of HPVAs and NHPVAs together revealed nearly significant associations between OS and HPV status and stage (both [P=0.06]). For DFS, stage was a significant variable (P=0.04), whereas HPV status and tumor size were nearly significant (P=0.06 and 0.07, respectively). Clinical outcome studies support the idea that the IECC classification not only separates ECAs on the basis of HPV status (usually assessed on H&E slides), but also has important clinical relevance.

Uterine perivascular epithelioid cell tumors (PEComas) are rare neoplasms that may show overlapping morphology and immunohistochemistry with uterine smooth muscle tumors. In this study, we evaluated the morphologic, immunohistochemical, and molecular features of 32 PEComas, including 11 with aggressive behavior. Two distinct morphologies were observed: classic (n=30) and those with a lymphangioleiomyomatosis appearance (n=2). In the former, patients ranged from 32 to 77 (mean: 51) years and 13% had tuberous sclerosis. Tumors ranged from 0.2 to 17 (mean: 5.5) cm with 77% arising in the corpus. Epithelioid cells were present in 100% and a spindled component was seen in 37%. Nuclear atypia was low (53%), intermediate (17%), or high (30%). Mitoses ranged from 0 to 36 (mean: 6) and 0 to 133 (mean: 19) per 10 and 50 high-power fields, with atypical mitoses present in 30%. Thin and delicate vessels were noted in 100%, clear/eosinophilic and granular cytoplasm in 93%, stromal hyalinization in 73%, necrosis in 30%, and lymphovascular invasion in 10%. All tumors were positive for HMB-45, cathepsin K, and at least one muscle marker, with most expressing melan-A (77%) and/or MiTF (79%). A PSF-TFE3 fusion was identified in one while another showed a RAD51B-OPHN1 fusion. Follow-up ranged from 2 to 175 (mean: 41) months, with 63% of patients alive and well, 20% dead of disease, 13% alive with disease, and 3% dead from other causes. In the latter group (n=2), patients were 39 and 49 years old, one had tuberous sclerosis, while the other had pulmonary lymphangioleiomyomatosis. Both tumors expressed HMB-45, cathepsin K, and muscle markers, but lacked TFE3 and RAD51B rearrangements. The 2 patients are currently alive and well. Application of gynecologic-specific criteria (≥4 features required for malignancy: size ≥5 cm, high-grade atypia, mitoses >1/50 high-power fields, necrosis, and lymphovascular invasion) for predicting outcome misclassified 36% (4/11) of aggressive tumors; thus, a modified algorithm with a threshold of 3 of these features is recommended to classify a PEComa as malignant.

The International Endocervical Adenocarcinoma Criteria and Classification was developed to separate endocervical adenocarcinomas (ECAs) into 2 main categories on the basis of morphology such as human papilloma virus-associated (HPVA) and non-human papilloma virus-associated adenocarcinomas. We aimed to improve the diagnostic accuracy of International Endocervical Adenocarcinoma Criteria and Classification by performing a comprehensive immunohistochemical evaluation and constructing objective immunohistochemical-based algorithms for the classification of these tumors. Tissue microarrays were constructed from 297 of 409 cases used to develop the original classification. Immunostains included p16, p53, estrogen receptor (ER), progesterone receptor, androgen receptor, Vimentin, CK7, CK20, HER2, HIK1083, MUC6, CA-IX, SATB2, HNF-1beta, napsin A, PAX8, CDX2, GATA3, p63, p40, and TTF-1. High-risk human papilloma virus (HR-HPV) was detected by in situ hybridization (ISH) using probes against E6 and E7 mRNA expressed in 18 different virus types. Vimentin, ER, and progesterone receptor were expressed in a significant minority of ECAs, mostly HPVAs, limiting their use in differential diagnosis of endometrioid carcinoma when unaccompanied by HPV-ISH or p16. HR-HPV ISH had superior sensitivity, specificity, and negative and positive predictive values compared with p16, as published previously. HNF-1beta did not have the anticipated discriminatory power for clear cell carcinoma, nor did MUC6 or CA-IX for gastric-type carcinoma. HNF-1beta and napsin A were variably expressed in clear cell carcinoma, with HNF-1beta demonstrating less specificity, as it was ubiquitously expressed in gastric-type carcinoma and in the majority of HPV-associated mucinous (predominantly intestinal-type and invasive ECA resembling stratified mucin-producing intraepithelial lesion [iSMILE]) and usual-type carcinomas. HIK1083 was expressed in nearly half of gastric-type carcinomas, but not in the vast majority of other subtypes. GATA3 was positive in 10% of usual-type adenocarcinomas and in single examples of other subtypes. Rare gastric-type and HPVA mucinous carcinomas displayed HER2 overexpression. Androgen receptor was positive in 6% of usual-type adenocarcinomas. Aberrant p53 expression was found in only 3.6% of usual-type HPVA carcinomas, but it was more prevalent in mucinous (intestinal type and iSMILE) HPVAs and non-human papilloma virus-associates (particularly in gastric-type carcinoma, >50% of cases). The following diagnostic classification algorithms were developed with the above data. Carcinomas without overt cytoplasmic mucin (endometrioid, usual-type endocervical, clear cell, and mesonephric carcinomas) can be subclassified using HR-HPV ISH, ER, and GATA3, whereas carcinomas with easily appreciated cytoplasmic mucin (endometrioid carcinoma with mucinous features, HPVA mucinous, and gastric-type carcinomas) can be subclassified with HR-HPV ISH and ER.

We sought to classify endocervical adenocarcinomas (ECAs) based on morphologic features linked to etiology (ie, human papillomavirus [HPV] infection), unlike the World Health Organization 2014 classification. The International Endocervical Adenocarcinoma Criteria and Classification (IECC criteria), described herein, distinguishes between human papillomavirus-associated adenocarcinoma (HPVA), recognized by the presence of luminal mitoses and apoptosis seen at scanning magnification, and no or limited HPVA features (nonhuman papillomavirus-associated adenocarcinoma [NHPVA]). HPVAs were then subcategorized based on cytoplasmic features (mostly to provide continuity with preexisting classification schemes), whereas NHPVAs were subclassified based on established criteria (ie, gastric-type, clear cell, etc.). Complete slide sets from 409 cases were collected from 7 institutions worldwide. Tissue microarrays representing 297 cases were constructed; immunohistochemistry (p16, p53, vimentin, progesterone receptor) and chromogenic in situ hybridization using an RNA-based probe set that recognizes 18 varieties of high-risk HPV were performed to validate IECC diagnoses. The 5 most common IECC diagnoses were usual-type (HPVA) (73% of cohort), gastric-type (NHPVA) (10%), mucinous adenocarcinoma of HPVA type, including intestinal, mucinous not otherwise specified, signet-ring, and invasive stratified mucin-producing carcinoma categories (9%), clear cell carcinoma (NHPVA) (3%) and adenocarcinoma, not otherwise specified (2%). Only 3 endometrioid carcinomas were recognized and all were NHPVA. When excluding cases thought to have suboptimal tissue processing, 90% and 95% of usual-type IECC cases overexpressed p16 and were HPV, whereas 37% and 3% of NHPVAs were p16 and HPV, respectively. The 1 HPV gastric-type carcinoma was found to have hybrid HPVA/NHPVA features on secondary review. NHPVA tumors were larger and occurred in significantly older patients, compared with HPVA tumors (P<0.001). The high-risk HPV chromogenic in situ hybridization probe set had superior sensitivity, specificity, and positive and negative predictive values (0.955, 0.968, 0.992, 0.833, respectively) compared with p16 immunohistochemistry (0.872, 0.632, 0.907, 0.545, respectively) to identify HPV-related usual carcinoma and mucinous carcinoma. IECC reliably segregates ECAs into HPVA and NHPVA types using morphology alone. This study confirms that usual-type ECAs are the most common type worldwide and that mucinous carcinomas comprise a mixture of HPVA and NHPVA, with gastric-type carcinoma being the major NHPVA type. Endometrioid and serous carcinomas of the endocervix are extraordinarily rare. Should clinical outcomes and genomic studies continue to support these findings, we recommend replacement of the World Health Organization 2014 criteria with the IECC 2017.

Adenocarcinomas exhibiting gastric differentiation represent a recently described and uncommon subtype of non-human papillomavirus (HPV)-related cervical adenocarcinoma. They comprise a spectrum from a well-differentiated variant (adenoma malignum/mucinous variant of minimal deviation adenocarcinoma) to a more poorly differentiated overtly malignant form, generally referred to as gastric-type adenocarcinoma. Rarely, such tumors have also been described as primary vaginal neoplasms. Gastric-type adenocarcinomas exhibit considerable morphologic overlap with adenocarcinomas originating outside the female genital tract, especially mucinous adenocarcinomas arising in the pancreas and biliary tract. Moreover, they often metastasize to unusual sites, such as the ovary and peritoneum/omentum, where they can be mistaken for metastatic adenocarcinomas from other, nongynecologic sites. There is little information regarding the immunophenotype of gastric-type adenocarcinomas, and knowledge of this is important to aid in the distinction from other adenocarcinomas. In this study, we undertook a detailed immunohistochemical analysis of a large series of cervical (n=45) and vaginal (n=2) gastric-type adenocarcinomas. Markers included were cytokeratin (CK)7, CK20, CDX2, carcinoembryonic antigen, CA125, CA19.9, p16, estrogen receptor, progesterone receptor, MUC6, PAX8, PAX2, p53, hepatocyte nuclear factor 1 beta, carbonic anhydrase IX, human epidermal receptor 2 (HER2), and mismatch repair (MMR) proteins. All markers were classified as negative, focal (<50% of tumor cells positive), or diffuse (≥50% tumor cells positive) except for p53 (classified as "wild-type" or "mutation-type"), HER2 (scored using the College of American Pathologists guidelines for gastric carcinomas), and MMR proteins (categorized as retained or lost). There was positive staining with CK7 (47/47-45 diffuse, 2 focal), MUC6 (17/21-6 diffuse, 11 focal), carcinoembryonic antigen (25/31-12 diffuse, 13 focal), carbonic anhydrase IX (20/24-8 diffuse, 12 focal), PAX8 (32/47-20 diffuse, 12 focal), CA125 (36/45-5 diffuse, 31 focal), CA19.9 (11/11-8 diffuse, 3 focal), hepatocyte nuclear factor 1 beta (13/14-12 diffuse, 1 focal), CDX2 (24/47-4 diffuse, 20 focal), CK20 (23/47-6 diffuse, 17 focal), and p16 (18/47-4 diffuse, 14 focal). Most cases were negative with estrogen receptor (29/31), progesterone receptor (10/11), PAX2 (18/19), and HER2 (25/26). p53 showed "wild-type" and "mutation-type" staining in 27 of 46 and 19 of 46 cases, respectively. MMR protein expression was retained in 19 of 20 cases with loss of MSH6 staining in 1 patient with Lynch syndrome. Molecular studies for HPV were undertaken in 2 tumors, which exhibited diffuse "block-type" immunoreactivity with p16, and both were negative. This is the first detailed immunohistochemical study of a large series of gastric-type adenocarcinomas of the lower female genital tract. Our results indicate immunophenotypic overlap with pancreaticobiliary adenocarcinomas but suggest that PAX8 immunoreactivity may be especially useful in distinguishing gastric-type adenocarcinomas from pancreaticobiliary and other nongynecologic adenocarcinomas, which are usually negative. Diffuse "block-type" p16 immunoreactivity in a cervical adenocarcinoma is not necessarily indicative of a high-risk HPV-associated tumor.

Gastric-type adenocarcinoma of the uterine cervix (GAS) is a rare variant of mucinous endocervical adenocarcinoma not etiologically associated with human papillomavirus (HPV) infection, with minimal deviation adenocarcinoma (MDA) at the well-differentiated end of the morphologic spectrum. These tumors are reported to have worse prognosis than usual HPV associated endocervical adenocarcinoma (UEA). A retrospective review of GAS was performed from the pathology databases of 3 institutions spanning 20 years. Stage, metastatic patterns, and overall survival were documented. Forty GAS cases were identified, with clinical follow-up data available for 38. The tumors were subclassified as MDA (n=13) and non-MDA GAS (n=27). Two patients were syndromic (1 Li-Fraumeni, 1 Peutz-Jeghers). At presentation, 59% were advanced stage (FIGO II to IV), 50% had lymph node metastases, 35% had ovarian involvement, 20% had abdominal disease, 39% had at least 1 site of metastasis at the time of initial surgery, and 12% of patients experienced distant recurrence. The metastatic sites included lymph nodes, adnexa, omentum, bowel, peritoneum, diaphragm, abdominal wall, bladder, vagina, appendix, and brain. Follow-up ranged from 1.4 to 136.0 months (mean, 33.9 mo); 20/38 (52.6%) had no evidence of disease, 3/38 (7.9%) were alive with disease, and 15/38 (39.5%) died of disease. Disease-specific survival at 5 years was 42% for GAS versus 91% for UEA. There were no survival differences between MDA and non-MDA GAS. GAS represents a distinct, biologically aggressive type of endocervical adenocarcinoma. The majority of patients present at advanced stage and pelvic, abdominal, and distant metastases are not uncommon.

Accurate distinction of clear cell carcinoma (CCC) from endometrioid carcinoma (EC) has important clinical implications, but, not infrequently, EC demonstrates clear cell change (EC-CC), mimicking CCC. We examined whether a panel of immunomarkers can help distinguish between these tumors. Sixty-four CCCs (40 ovarian and 24 uterine), 34 ECs (21 ovarian and 13 uterine), and 34 EC-CCs (6 ovarian and 28 uterine) were stained for HNF1β, BAF250a, Napsin A, ER, and PR. Intensity and extent of immunoreactivity was assessed. Fifty-seven of 64 (89%) CCCs, 14/34 (41%) EC-CCs, and 16/34 (47%) ECs expressed HNF1β, and 56/64 (88%) CCCs, 4/34 (12%) EC-CCs, and 1/34 (3%) ECs stained for Napsin A. Most CCCs demonstrated at least moderate and diffuse staining for both markers, whereas only focal and weak expression was identified in most EC-CC/EC. Compared to HNF1β, Napsin A showed increased specificity (93.0% vs. 55.9%, P<0.0001) and similar sensitivity (87.5% vs. 89.1%) in distinguishing CCC from EC-CC/EC. Thirteen of 64 (20%) CCCs, 6/34 (18%) EC-CCs, and 2/34 (6%) ECs showed loss of BAF250a. ER was expressed by 10/64 (16%) CCCs, 30/34 (88%) EC-CCs, and 33/34 (97%) ECs, whereas PR positivity was identified in 9/64 (14%) CCCs, 26/34 (77%) EC-CCs, and 33/34 (97%) ECs. The majority of EC and EC-CC demonstrated diffuse staining for ER/PR, whereas most CCCs showed very focal positivity. There is a statistically significant difference in HNF1β, Napsin A, ER, and PR immunoexpression between CCC and EC/EC-CC, with Napsin A being a more specific marker for CCC than HNF1β. Overall, the immunoprofile of EC-CC is more comparable to that of EC than CCC. The use of a panel of immunostains can help distinguish EC-CC from CCC.

Background: Endocervical adenocarcinomas of the usual type are etiologically related to infection with oncogenic human papillomaviruses (HPVs). These tumors are typically diffusely positive for p16 and carcinoembryonic antigen (CEA) immunostains. The goal of our study was to determine the HPV status and immunohistochemical profiles of unusual histologic subtypes of endocervical adenocarcinoma., Methods: The study consisted of a total of 26 cases of unusual subtypes including clear cell carcinoma (CCC, n=9), gastric-type adenocarcinoma (GAS, n=11), minimal deviation adenocarcinoma (MDA, n=3), mesonephric adenocarcinoma (MSN, n=1), serous adenocarcinoma (SER, n=1), and malignant mixed Müllerian tumor (n=1). In addition, 5 cases of usual-type endocervical adenocarcinoma (UEA) were included in the study as a control group. The cases were tested for HPV using SPF-10 PCR and LiPA assays, and immunostained for p16, HIK1083, hepatocyte nuclear factor 1-β, p53, CEA, estrogen receptor (ER), and progesterone receptor (PR)., Results: HPV DNA was not detected in any of the unusual adenocarcinoma subtypes, with the exception of a single case of SER in which HPV16 was detected. p16 positivity did not correlate with HPV status, as 42% of HPV-negative tumors showed patchy/diffuse p16 overexpression; however, p16 positivity was uncommon in GAS/MDA. HIK1083 positivity was limited to GAS and MDA, indicating relative specificity for tumors with gastric mucin expression. Hepatocyte nuclear factor 1-β was positive in the majority of CCCs and also in other tumor variants and in some UEA as well, indicating a lack of specificity for clear cell differentiation. CEA was consistently negative in CCCs and in a single MSN, but positive in GAS, MDA, SER, and UEA, suggesting that it may serve as a negative marker of clear cell differentiation. p53 was diffusely positive in almost half of the GAS cases, whereas UEA showed mostly negative staining and other variants showed focal staining. PR was negative in all variant cases and in all UEA. ER expression, although mostly negative, showed focal staining in a few variant cases and UEA., Conclusions: Unusual variants of endocervical adenocarcinoma are not related to HPV infection, with only rare exceptions, and p16 overexpression in non-UEA does not correlate with HPV status. Negative staining for PR and ER may serve as a general marker of endocervical neoplasia. GAS/MDA may be differentiated from all other adenocarcinomas with either positive HIK1083 stain or negative/focal p16 stain. Positive CEA stain differentiates GAS/MDA from CCC and negative PR and ER stains differentiate GAS/MDA from benign endocervical glands. CCC may be distinguished from all other adenocarcinomas, except MSN, with a negative CEA stain. Strong and diffuse p53 positivity in SER may be useful in differentiation from UEA. MSN may be identified with negative CEA, ER, and PR stains.

120 Krukenberg tumors were analyzed with emphasis on their wide microscopic spectrum and resultant problems in differential diagnosis. The patients ranged from 13 to 84 years (average, 45 years) with 43% of them under 40 years. Abdominal swelling or pain usually accounted for the clinical presentation, but 17 had abnormal vaginal bleeding, 4 had virilization, and 4 had hirsutism without virilization. Ascites was present in 43% of the cases. Sixty-three percent of the tumors were documented to be bilateral, but both ovaries were not always removed or rigorously examined microscopically. The mean diameter of the tumors was 10.4 cm, and they typically had intact, bosselated external surfaces without adhesions. The sectioned surfaces were typically solid and firm to edematous to gelatinous; one third of the tumors also had cysts. Microscopic examination showed great variation from case to case and within individual neoplasms. Multiple nodules separated by normal stroma were seen in small neoplasms and focally in many larger ones. The tumors were often more cellular at their periphery and edematous to gelatinous centrally. An irregular distribution of cellular and less cellular areas often imparted a pseudolobular pattern. The cellularity of the stroma ranged from densely cellular to paucicellular; the latter regions ranged from edematous to mucoid. The overall morphology varied according to the prominence of signet-ring cells, extracellular mucin, edema, and various epithelial patterns. Signet-ring cells were numerous in most neoplasms (and by definition occupied at least 10% of the neoplasm) but were often absent or inconspicuous in significant areas of them. The signet-ring cells typically had modest but sometimes copious amounts of pale to basophilic cytoplasm; occasionally, it was eosinophilic. The signet-ring cells varied widely in their arrangement, growing singly, in clusters, forming confluent masses or pseudo-tubular arrays or lining part of all of a true tubule. Small glands and tubules were common, often resembling microcysts (when the lining cells were flattened) or Sertoli tubules; mucinous glands and cysts and medium-sized to large intestinal-type glands were also relatively common, particularly the latter. Extracellular mucin was often conspicuous and, when associated with scant acellular collagenous stroma, gave a distinctive appearance referred to by us as "feathery degeneration." Stromal luteinization was present in the tumors of the 8 pregnant patients and was seen in 14% of the nonpregnant patients. Unusual features that complicated the microscopic picture included diffuse sheets or other arrangements of mucin-free indifferent cells, squamous cells, clear cells, transitional cells, and corded, trabecular, and insular patterns. Vascular space invasion was common. Two thirds of the primary carcinomas were detected synchronously with, or subsequent to, detection of the Krukenberg tumor compounding the diagnostic difficulty posed by the cases. Two thirds of the primary tumors were in the stomach; other primary sites in order of frequency were appendix, colon, breast, small intestine, rectum, gallbladder, and urinary bladder. Our observations emphasize that the microscopic spectrum of the Krukenberg tumor is broader that often presented in the literature, in particular tubules, glands, and cysts often being present, and the wide pathologic differential diagnosis is discussed.